Title: Research on interval triangular fuzzy multi-attribute fault diagnosis methods based on the grey relation grade
Authors: Yin Liu; Zhaoqin Lu
Addresses: Mechanical and Electronic Engineering College, Shandong Agricultural University, Taian 271018, China; School of Mechanical and Automotive Engineering, Qilu University of Technology, Jinan 250353, China ' Mechanical and Electronic Engineering College, Shandong Agricultural University, Taian 271018, China
Abstract: Considering the fuzzy multi-attribute problems of components' failure information, a fault diagnosis model of a Bayesian network based on the interval triangular fuzzy multi-attribute decision method was built. First, the fault multi-attribute hierarchies were represented as a Bayesian network structure by Bayesian network causal dependencies, which reduced the complexity of the multi-attribute hierarchy. Second, in the condition that weight information is completely unknown, grey correlation optimal models were constructed based on the grey correlation analysis method. Furthermore, faulty components were sorted and the component with the highest value was optimised, and the calculation method of interval triangle fuzzy multi-attribute decision based on the grey correlation was given, which realised the fault diagnosis of the system fuzzy multiple-attribute problem. Finally, the method was applied in an analysis of a case of fault diagnosis on a CNC machine tool servo system with a low voltage alarm, which verified the effectiveness of the method.
Keywords: grey correlation degree; interval triangle; multiple-attribute decision; Bayesian network; fault diagnosis.
International Journal of Industrial and Systems Engineering, 2018 Vol.28 No.3, pp.331 - 345
Received: 29 Aug 2015
Accepted: 05 Feb 2016
Published online: 09 Feb 2018 *